Method, apparatus, electronic device and storage medium for constructing a real-scene three-dimensional model

Abstract

The present invention discloses a method, apparatus, electronic device and storage medium for constructing a real-scene three-dimensional model. The present invention uses digital elevation data to generate a triangular network terrain three-dimensional surface of a target to be constructed, and geometrically corrects the surface, so as to use the corrected surface to reconstruct the geometric structure of an initial three-dimensional real-scene model obtained by point cloud reverse modeling technology. Thus, the present invention is equivalent to constructing a real-scene three-dimensional model with high spatial accuracy by using point cloud reverse modeling technology and geometric terrain surface correction technology. Therefore, the present invention effectively solves the problem of ribbon holes or independent holes existing in the real-scene three-dimensional model obtained by point cloud reverse modeling technology, making the reconstructed real-scene three-dimensional model have a high utilization rate of previous surveying and mapping results, and having the advantages of fast and simple operation, no holes in the model, high accuracy, high spatial resolution, good texture mapping effect and multiple result data formats, and is suitable for large-scale application and promotion.

Description

Technical Field

[0001] The present invention belongs to the technical field of three-dimensional real-scene modeling, and particularly relates to a method, device, electronic device and storage medium for constructing an actual three-dimensional model. Background Art

[0002] The main form of real-scene three-dimensional is oblique photography three-dimensional modeling. This measurement technology acquires high-resolution images from five different angles, and then constructs a high-precision and high-spatial-resolution urban-level three-dimensional real-scene model based on the acquired high-resolution images. However, this method relies on external data collection, has high requirements for drone equipment, and the amount of data obtained is large, and the model reconstruction cycle is long. In addition, for terrain and features with low reflectivity, the modeling effect is poor and further model modification is required, resulting in an increase in time and labor costs. The terrain-level three-dimensional real-scene model can not only use existing spatial data bodies (such as geographical scene DEM (Digital Elevation Model) data and DOM (Digital Orthophoto Map) data) as data sources, but also quickly and conveniently generate a three-dimensional real-scene model that meets the accuracy requirements. Therefore, the terrain-level three-dimensional real-scene model is gradually applied to the macro planning of various industries.

[0003] Currently, there are the following three methods for terrain-level three-dimensional real-scene modeling: (1) The three-dimensional analysis and modeling method based on software such as ArcGIS and MapGIS. This method can realize the fusion and superposition rendering of DEM and DOM data to obtain a three-dimensional real-scene terrain. Its operation is simple, but it is limited to viewing and analyzing in the above software and cannot be stored in other data formats, resulting in limited application scope. (2) The three-dimensional modeling method based on software such as Aerospace Vision and SuperMap. This method can directly generate a real-scene three-dimensional model based on DEM and DOM data. However, the MM3D module of Aerospace Vision can only generate a non-standard osgb format model, and the fbx, osgb and other format models obtained by SuperMap software have no real coordinate information and are only for browsing and display. Therefore, the application scope of method (2) is also very limited. (3) Using DEM data to obtain point cloud data and reconstructing a model based on the point cloud data to obtain a three-dimensional real-scene model. This method has high modeling efficiency, a variety of output data formats, and the model accuracy can reach the same accuracy level as DEM and DOM. Therefore, it has broad application prospects. However, the three-dimensional real-scene model reconstructed by this method is prone to ribbon-shaped holes or independent holes, resulting in an incomplete model and poor application effect. In view of this, how to eliminate the influence of holes on the model to obtain a three-dimensional real-scene model with more complete texture, richer formats and real coordinate information has become an urgent problem to be solved. Summary of the Invention

[0004] The object of the present invention is to provide a method, apparatus, electronic device and storage medium for constructing a real-scene three-dimensional model, so as to solve the problems in the prior art that when reconstructing a model based on point cloud data, ribbon-shaped holes or independent holes are likely to appear in the model, resulting in an incomplete model and poor application effects.

[0005] To achieve the above object, the present invention adopts the following technical solutions:

[0006] In a first aspect, a method for constructing a real-scene three-dimensional model is provided, including:

[0007] Obtain model construction data of a target to be constructed, where the model construction data includes digital elevation data and digital orthophoto image data of the target to be constructed;

[0008] Based on the digital elevation data, determine point cloud data for reconstructing the three-dimensional model of the target to be constructed, and use the digital orthophoto image data to perform point cloud coloring processing on the point cloud data to obtain colored point cloud data;

[0009] Based on the colored point cloud data, construct an initial real-scene three-dimensional model of the target to be constructed;

[0010] Use the digital elevation data to construct a triangular network terrain three-dimensional surface of the target to be constructed;

[0011] Perform geometric correction processing on the triangular network terrain three-dimensional surface to obtain a corrected triangular network terrain three-dimensional surface;

[0012] Use the corrected triangular network terrain three-dimensional surface to perform geometric structure reconstruction processing on the initial real-scene three-dimensional model, so as to obtain the real-scene three-dimensional model of the target to be constructed after the geometric structure reconstruction processing.

[0013] Based on the above - disclosed content, the present invention first uses digital elevation data to obtain the point - cloud data for model reconstruction, and then, with the aid of the color information in the digital orthophoto image data, performs coloring processing on the obtained point - cloud data to obtain colored point - cloud data. Then, the point - cloud reverse modeling technology and the colored point - cloud data can be used to realize the reconstruction of the three - dimensional real - scene model of the target to be constructed, and an initial three - dimensional real - scene model is obtained. At the same time, to avoid the existence of ribbon holes or isolated holes in the reconstructed model, the present invention also uses digital elevation data to reconstruct the triangular - mesh terrain three - dimensional surface of the target to be constructed, and through geometric correction of the triangular - mesh terrain three - dimensional surface, a corrected triangular - mesh terrain three - dimensional surface is obtained. In this way, this step is equivalent to obtaining the true three - dimensional geometric structure of the target to be constructed. Finally, the corrected triangular - mesh terrain three - dimensional surface can be used to perform geometric structure reconstruction processing on the previously reconstructed initial three - dimensional real - scene model, thereby completing the geometric structure correction of the initial three - dimensional real - scene model to remove the ribbon holes and isolated holes in the model, and further improving the integrity, accuracy, spatial resolution, and texture mapping effect of the model.

[0014] Through the above design, the present invention uses digital elevation data to generate the triangular - mesh terrain three - dimensional surface of the target to be constructed, and performs geometric correction on this surface, so as to use the corrected surface to perform geometric structure reconstruction on the initial three - dimensional real - scene model obtained by the point - cloud reverse modeling technology. In this way, the present invention is equivalent to using the point - cloud reverse modeling technology and the geometric terrain surface correction technology to construct a real - scene three - dimensional model with high spatial accuracy. Therefore, the present invention effectively solves the problem of ribbon holes or isolated holes existing in the real - scene three - dimensional model obtained by the point - cloud reverse modeling technology, making the reconstructed real - scene three - dimensional model have a high utilization rate of the previous surveying and mapping results, and having the advantages of fast and simple operation, no holes in the model, high accuracy, high spatial resolution, good texture mapping effect, and multiple output data formats, and is suitable for large - scale application and promotion.

[0015] In a possible design, based on the digital elevation data, determining the point - cloud data for reconstructing the three - dimensional model of the target to be constructed includes:

[0016] Using the bilinear interpolation algorithm to perform bilinear interpolation processing on the digital elevation data to obtain an interpolated digital elevation data set;

[0017] Determining the grid - shaped point - cloud data with three - dimensional coordinate information from the interpolated digital elevation data set, and using the grid - shaped point - cloud data as the point - cloud data for reconstructing the three - dimensional model of the target to be constructed.

[0018] Based on the above - disclosed content, the present invention discloses the specific process of obtaining point - cloud data. First, perform interpolation processing on the digital elevation data to narrow the grid spacing between each data in the digital elevation data, so as to obtain a larger number of digital elevation data. In this way, the density of the extracted point - cloud data can be increased, and thus fine textures can be obtained. Then, extract the grid - shaped point - cloud data with three - dimensional coordinate information from the interpolated digital elevation data set. Finally, the extracted grid - shaped point - cloud data can be used as the point - cloud data for reconstructing the three - dimensional model of the target to be constructed. Through the above design, the present invention can increase the point - cloud density for model reconstruction, thereby maximizing the accuracy of the model.

[0019] In a possible design, determining the grid - shaped point - cloud data with three - dimensional coordinate information from the interpolated digital elevation data set includes:

[0020] Call the pixel - center - point extraction component in Global Mapper to perform grid - point - cloud extraction processing on the interpolated digital elevation data set, so as to obtain the grid - shaped point - cloud data with three - dimensional coordinate information after the grid - point - cloud extraction processing.

[0021] In a possible design, performing point - cloud coloring processing on the point - cloud data by using the digital orthophoto image data to obtain colored point - cloud data includes:

[0022] Call the pixel - matching component in Global Mapper to use the pixel - matching component to match each pixel data in the digital orthophoto image data with the point - cloud data, so as to determine the pixel data matching each point - cloud data in the digital orthophoto image data;

[0023] For any point - cloud data, obtain the color information of the pixel data corresponding to the any point - cloud data as the coloring information;

[0024] Add the coloring information to the any point - cloud data to complete the point - cloud coloring processing of the any point - cloud data after the addition.

[0025] Based on the above - disclosed content, the present invention discloses the specific process of performing coloring processing on point - cloud data. Among them, since the spatial positions of the digital elevation data and the digital orthophoto image data are in one - to - one correspondence, therefore, first, with the help of the pixel - matching component in Global Mapper, determine the pixel data matching each point - cloud data (that is, representing the same position) from the digital orthophoto image data. Then, add the color information in the pixel to its corresponding point - cloud data to obtain real - colored point - cloud data. In this way, the coloring processing of each point - cloud data can be completed, and thus colored point - cloud data can be obtained.

[0026] In a possible design, geometric correction processing is performed on the triangular mesh terrain three-dimensional surface to obtain a corrected triangular mesh terrain three-dimensional surface, including:

[0027] The triangular mesh terrain three-dimensional surface is subjected to clipping processing to obtain a number of clipped terrain three-dimensional surfaces;

[0028] Anti-aliasing processing is performed on each of the number of clipped terrain three-dimensional surfaces to obtain a number of anti-aliased clipped terrain three-dimensional surfaces;

[0029] The number of anti-aliased clipped terrain three-dimensional surfaces are subjected to merging processing to obtain a preprocessed triangular mesh terrain three-dimensional surface;

[0030] The preprocessed triangular mesh terrain three-dimensional surface is subjected to smoothing processing to obtain the corrected triangular mesh terrain three-dimensional surface after the smoothing processing.

[0031] Based on the above-disclosed content, the present invention discloses the specific process of geometric correction processing for the triangular mesh terrain three-dimensional surface, that is, first clipping the three-dimensional surface to obtain a number of clipped three-dimensional surfaces; then, removing the jaggedness in the clipped three-dimensional surfaces to obtain anti-aliased clipped three-dimensional surfaces; finally, the three-dimensional surfaces processed as described above can be merged and surface smoothing processing can be performed, thereby solving the problems of non-smoothness and irregularity existing on the surface of the terrain surface; thus, the present invention can achieve surface repair of the surface model of the to-be-constructed target to obtain a real three-dimensional geometric structure, so as to provide a high-precision data basis for subsequent model correction of the initial real-scene three-dimensional model.

[0032] In a possible design, the preprocessed triangular mesh terrain three-dimensional surface is subjected to smoothing processing to obtain the corrected triangular mesh terrain three-dimensional surface after the smoothing processing, including:

[0033] Obtain a smoothing angle, and based on the smoothing angle, perform smoothing processing on the preprocessed triangular mesh terrain three-dimensional surface to obtain the corrected triangular mesh terrain three-dimensional surface.

[0034] In a possible design, using the corrected triangular mesh terrain three-dimensional surface, geometric structure reconstruction processing is performed on the initial real-scene three-dimensional model, including:

[0035] Based on the corrected triangular mesh terrain three-dimensional surface, obtain three-dimensional geometric structure data of the to-be-constructed target;

[0036] Using the three-dimensional geometric structure data, adjust the geometric structure of the initial real-scene three-dimensional model to obtain a preprocessed real-scene three-dimensional model of the to-be-constructed target after the adjustment;

[0037] Perform texture replacement processing on the preprocessed real-scene three-dimensional model, so as to obtain the real-scene three-dimensional model of the target to be constructed after the texture replacement processing.

[0038] Based on the above-disclosed content, the present invention is equivalent to using the corrected triangular network terrain three-dimensional surface to obtain the real three-dimensional geometric structure of the target to be constructed. Then, the three-dimensional geometric structure can be used to perform geometric correction on the initial real-scene three-dimensional model; finally, texture replacement is performed. After the replacement is completed, a real-scene three-dimensional model without ribbon holes and independent holes can be obtained.

[0039] In a second aspect, a device for constructing a real-scene three-dimensional model is provided, including:

[0040] A data acquisition unit, configured to acquire model construction data of the target to be constructed, where the model construction data includes digital elevation data and digital orthophoto image data of the target to be constructed;

[0041] A point cloud processing unit, configured to determine point cloud data for reconstructing the three-dimensional model of the target to be constructed based on the digital elevation data, and perform point cloud coloring processing on the point cloud data by using the digital orthophoto image data to obtain colored point cloud data;

[0042] A model reconstruction unit, configured to construct an initial real-scene three-dimensional model of the target to be constructed based on the colored point cloud data;

[0043] A surface construction unit, configured to construct a triangular network terrain three-dimensional surface of the target to be constructed by using the digital elevation data;

[0044] A geometric correction processing unit, configured to perform geometric correction processing on the triangular network terrain three-dimensional surface to obtain a corrected triangular network terrain three-dimensional surface;

[0045] A model adjustment unit, configured to perform geometric structure reconstruction processing on the initial real-scene three-dimensional model by using the corrected triangular network terrain three-dimensional surface, so as to obtain the real-scene three-dimensional model of the target to be constructed after the geometric structure reconstruction processing.

[0046] In a third aspect, another device for constructing a real-scene three-dimensional model is provided. Taking the device as an electronic device as an example, it includes a memory, a processor, and a transceiver that are communicatively connected in sequence, where the memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program and execute the method for constructing the real-scene three-dimensional model as described in the first aspect or any one of the possible designs in the first aspect.

[0047] Fourthly, a storage medium is provided, on which instructions are stored. When the instructions run on a computer, the method for constructing the real-scene three-dimensional model as described in the first aspect or any possible design in the first aspect is executed.

[0048] Fifthly, a computer program product containing instructions is provided. When the instructions run on a computer, the computer is made to execute the method for constructing the real-scene three-dimensional model as described in the first aspect or any possible design in the first aspect.

[0049] Beneficial effects:

[0050] (1) The present invention uses digital elevation data to generate a triangular network terrain three-dimensional surface of the target to be constructed, and geometrically corrects the surface, so as to use the corrected surface to reconstruct the geometric structure of the initial three-dimensional real-scene model obtained by the point cloud reverse modeling technology. In this way, the present invention is equivalent to using the point cloud reverse modeling technology and the geometric terrain surface correction technology to construct a real-scene three-dimensional model with high spatial accuracy. Therefore, the present invention effectively solves the problems of ribbon holes or independent holes existing in the real-scene three-dimensional model obtained by the point cloud reverse modeling technology, making the reconstructed real-scene three-dimensional model have a high utilization rate of the previous surveying and mapping results, and having the advantages of fast and simple operation, no holes in the model, high accuracy, high spatial resolution, good texture mapping effect and multiple result data formats, and is suitable for large-scale application and promotion.

[0051] (2) The present invention is not only applicable to the scenario of constructing a terrain-level real-scene three-dimensional model based on the fusion of DEM and DOM data, but also effective in dealing with the holes in the model built from the airborne lidar point cloud data of drones. The reason is that compared with the point cloud obtained by three-dimensional laser scanning, the point cloud density and accuracy of the airborne lidar point cloud data of drones are both lower, and the three-dimensional model will also have ribbon-shaped holes and independent holes. By correcting the geometric structure of the model, the effect of patching holes in the model can be achieved. Description of the drawings

[0052] Figure 1 It is a schematic flow chart of the steps of the method for constructing the real-scene three-dimensional model provided by the embodiment of the present invention;

[0053] Figure 2 It is an extraction effect diagram of the point cloud after bilinear interpolation provided by the embodiment of the present invention;

[0054] Figure 3 It is an effect diagram of the colored point cloud data provided by the embodiment of the present invention;

[0055] Figure 4 It is an effect schematic diagram of the initial real-scene three-dimensional model of Yanmengou in area A provided by the embodiment of the present invention;

[0056] Figure 5Schematic diagram of the effect of the three-dimensional curved surface of the triangulation terrain after correction of Yanmengou in Area A provided by the embodiment of the present invention;

[0057] Figure 6 Schematic diagram of the effect of the real three-dimensional model of Yanmengou in Area A provided by the embodiment of the present invention;

[0058] Figure 7 Schematic diagram of the structure of the construction device of the real three-dimensional model provided by the embodiment of the present invention;

[0059] Figure 8 Schematic diagram of the structure of the electronic device provided by the embodiment of the present invention. Detailed implementation manners

[0060] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the present invention in combination with the accompanying drawings and the descriptions of the embodiments or the prior art. Obviously, the following descriptions of the structures of the drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts. It should be noted here that the descriptions of these embodiments are used to help understand the present invention, but do not constitute a limitation to the present invention.

[0061] It should be understood that although terms such as first and second may be used herein to describe various units, these units should not be limited by these terms. These terms are only used to distinguish one unit from another. For example, the first unit can be called the second unit, and similarly the second unit can be called the first unit, without departing from the scope of the exemplary embodiments of the present invention.

[0062] It should be understood that for the term "and / or" that may appear in this article, it is only a description of the association relationship of associated objects, indicating that three relationships can exist. For example, A and / or B can represent: A exists alone, B exists alone, and both A and B exist at the same time; for the term " / and" that may appear in this article, it is a description of another association object relationship, indicating that two relationships can exist. For example, A / and B can represent: A exists alone, and both A and B exist; in addition, for the character " / " that may appear in this article, generally it means that the front and rear associated objects are in an "or" relationship.

[0063] Embodiment:

[0064] See Figure 1As shown, the method for constructing a real - scene three - dimensional model provided in this embodiment uses digital elevation data to generate a triangular - mesh terrain three - dimensional surface of the target to be constructed, and geometrically corrects this surface, so as to use the corrected surface to reconstruct the geometric structure of the initial three - dimensional real - scene model obtained by the point - cloud reverse modeling technology. In this way, the present invention can effectively solve the problems of ribbon holes and independent holes existing in the real - scene three - dimensional model obtained by using traditional technologies, thereby improving the integrity, accuracy, and texture effect of the model; in this embodiment, for example, this method can run at the real - scene three - dimensional model reconstruction end - side. Among them, for example, the real - scene three - dimensional model reconstruction end can be, but is not limited to, a personal computer (PC) or a tablet computer. It can be understood that the foregoing execution subject does not constitute a limitation on the embodiments of the present application. Correspondingly, the running steps of this method can be, but are not limited to, as shown in the following steps S1 - S6.

[0065] S1. Obtain the model construction data of the target to be constructed. Among them, the model construction data includes the digital elevation data and digital orthophoto image data of the target to be constructed; in specific applications, for example, the data elevation data of the target to be constructed can be, but is not limited to, the national standard data with a resolution of 1:10,000, or the digital elevation data of the target to be constructed can be extracted from the digital line - drawn map with a resolution of 1:10,000 and a scale of 1:2,000; while the digital orthophoto image data of the target to be constructed is extracted from the orthophoto remote - sensing image map of the third national land survey and the river - lake demarcation data; of course, the user can also collect the digital orthophoto image data of the target to be constructed by himself; in this implementation, both the digital elevation data and the digital orthophoto image data are data collected under the 2000 National Geodetic Coordinate System (CGCS2000). Among them, the digital elevation data is specifically the data elevation data with a 1:10,000 grid and a spacing of 2.5 meters, and its data format can be, but is not limited to, the GRID format; while the digital orthophoto image data actually contains the orthophoto image of the target to be constructed and the spatial information of this orthophoto image (that is, the three - dimensional coordinate information and color information, etc. of each pixel point in the image). Optionally, for example, the orthophoto image is in the TIFF format, and the spatial information is in the ASCII format; in addition, for example, the target to be constructed can be, but is not limited to, including: basins, mountains, and hills, etc. Of course, the target to be constructed can be specifically selected according to actual use, and is not limited to the foregoing examples here.

[0066] After obtaining the model construction data of the target to be constructed, the digital elevation data in the model construction data can be used to reconstruct the model first, and then the model construction data is used to geometrically correct the reconstructed model to remove the ribbon holes and independent holes in the model; among them, the process of reconstructing the model using the digital elevation data can be, but is not limited to, as shown in the following steps S2 and S3.

[0067] S2. Based on the digital elevation data, determine the point cloud data for reconstructing the 3D model of the target to be constructed, and use the digital orthophoto data to perform point cloud coloring processing on the point cloud data to obtain colored point cloud data; specifically in application, for example, the determination process of the foregoing point cloud data may include but is not limited to the following steps S21 and S22.

[0068] S21. Use the bilinear interpolation algorithm to perform bilinear interpolation processing on the digital elevation data to obtain an interpolated digital elevation data set; in this embodiment, since the grid spacing of the foregoing digital elevation data is 2.5 m, the total number of grid point cloud data that can be extracted is small, and fine textures cannot be obtained; therefore, to improve the texture effect of model reconstruction, it is necessary to increase the amount of digital elevation data, thereby increasing the density of the grid point cloud data extracted from the data elevation data; specifically, in this embodiment, the bilinear interpolation algorithm is used to perform interpolation processing on the digital elevation data, thereby reducing its grid spacing, obtaining more point clouds, and the bilinear interpolation method performs linear interpolation in two directions, which can maximize the accuracy of the original data to further improve the texture accuracy.

[0069] After obtaining the interpolated digital elevation data set, the point cloud data can be extracted from this data set, as shown in the following step S22.

[0070] S22. Determine the grid point cloud data with three-dimensional coordinate information from the interpolated digital elevation data set, and use the grid point cloud data as the point cloud data for reconstructing the 3D model of the target to be constructed; specifically in implementation, for example, a Global Mapper software (a map drawing software) is preset in the real-scene 3D model reconstruction end. Therefore, the real-scene 3D model reconstruction end can call the pixel center point extraction component in Global Mapper to perform grid point cloud extraction processing on the interpolated digital elevation data set, so as to obtain the grid point cloud data with three-dimensional coordinate information after the grid point cloud extraction processing; after obtaining the point cloud data for establishing the real-scene 3D model of the foregoing target to be constructed, point cloud coloring processing can be performed, so as to use the colored point cloud data to construct the initial actual 3D model of the target to be constructed; among them, the point cloud coloring processing process may include but is not limited to the following steps S23 to S25.

[0071] S23. Call the pixel matching component in Global Mapper to use the pixel matching component to match each pixel data in the digital orthophoto data with the point cloud data, so as to determine the pixel data matching each point cloud data in the digital orthophoto data.

[0072] S24. For any point cloud data, obtain the color information of the pixel data corresponding to the any point cloud data as the coloring information.

[0073] S25. Add the coloring information to the any point cloud data to complete the point cloud coloring process after the addition.

[0074] The processing principles of the foregoing steps S23 to S25 are as follows: Since it has been previously described that the digital elevation data and the digital orthophoto image data are data in the same coordinate system, the spatial information of the two types of data is completely consistent. Therefore, in this embodiment, the pixel matching component in Global Mapper is first used to match the point cloud data corresponding to each pixel from the digital orthophoto image data; then, the color information corresponding to each pixel data is added to the point cloud data corresponding to its respective position, and the point cloud coloring process can be completed; after the foregoing processing, color information can be assigned to each point cloud data, enhancing the rendering of the real geographical environment. The addition of color information facilitates the identification and application of ground objects, obtaining a visualized effect of the real scene, thereby providing a real scene basis for the construction of the terrain-level three-dimensional real scene model.

[0075] After obtaining the colored point cloud data, the point cloud reverse modeling technology can be used to construct the initial real scene three-dimensional model of the target to be constructed, where the construction process is as shown in step S3 below.

[0076] S3. Based on the colored point cloud data, construct the initial real scene three-dimensional model of the target to be constructed; in specific applications, for example, the real scene three-dimensional model reconstruction terminal can but is not limited to pre-setting Smart 3D software. Therefore, Smart 3D software can be directly called to reconstruct the colored point cloud data. At the same time, during the reconstruction process, the point cloud attributes can be used as the color source for the point cloud three-dimensional display and texture selection of the reconstructed three-dimensional model. After the foregoing processing, a preliminary real scene three-dimensional model can be obtained; thus, through the foregoing design, reverse three-dimensional reconstruction using colored point cloud data is adopted. Since it does not require ground control points, aerial triangulation encryption, and image matching, it has strong robustness and higher 3D reconstruction processing efficiency, saving a large amount of modeling time; however, this model has defects such as ribbon holes and independent holes, making the model incomplete. Therefore, this embodiment also sets a model correction processing step to remove the ribbon holes and independent holes in the foregoing initial real scene three-dimensional model; optionally, the model correction process can but is not limited to the following steps S4 to S6.

[0077] S4. Using the digital elevation data, a triangular network terrain three-dimensional surface of the target to be constructed is constructed; in specific applications, for example, the real-scene three-dimensional model reconstruction terminal can but is not limited to presetting the 3ds Max software. Therefore, the digital elevation data can be directly imported into the 3ds Max software, and based on this 3ds Max software, the triangular network terrain three-dimensional surface of the target to be constructed can be generated. In this embodiment, the triangular network terrain three-dimensional surface is a three-dimensional geometric representation of the target to be constructed and can reflect its three-dimensional geometric structure. However, due to errors in surveying and mapping data and defects in the surface construction algorithm, the surface of the terrain three-dimensional surface is not smooth and irregular, resulting in holes in the triangular network terrain three-dimensional surface. Thus, it is necessary to perform geometric correction on this triangular network terrain three-dimensional surface, that is, to perform surface smoothing processing, so as to achieve the surface correction of the foregoing three-dimensional surface. The geometric correction process of the triangular network terrain three-dimensional surface can but is not limited to the steps shown in step S5 below.

[0078] S5. Perform geometric correction processing on the triangular network terrain three-dimensional surface to obtain a corrected triangular network terrain three-dimensional surface; in specific applications, it can but is not limited to sequentially performing cropping, anti-aliasing processing, and smoothing processing on the foregoing triangular network terrain three-dimensional surface. After the processing is completed, a corrected triangular network terrain three-dimensional surface can be obtained. Optionally, the foregoing process can but is not limited to the steps shown in steps S51 to S54 below.

[0079] S51. Perform cropping processing on the triangular network terrain three-dimensional surface to obtain a number of cropped terrain three-dimensional surfaces; in this embodiment, for example, it can but is not limited to calling the Global Mapper software to crop the foregoing triangular network terrain three-dimensional surface, so as to obtain a number of cropped terrain three-dimensional surfaces. Optionally, it can but is not limited to presetting different research areas of the target to be constructed. Therefore, the surface can be cropped according to the foregoing research areas. At the same time, for example, any cropped terrain three-dimensional surface can but is not limited to a file in the format of "*.dem". After the cropping of the triangular network terrain three-dimensional surface is completed, since there are still jagged edges in the surface image, it is also necessary to perform anti-aliasing processing on the cropped three-dimensional surface to reduce the interference of the jagged edges, as shown in step S52 below.

[0080] S52. Perform anti-aliasing processing on each cropped terrain three-dimensional surface among the number of cropped terrain three-dimensional surfaces to obtain a number of anti-aliased cropped terrain three-dimensional surfaces; in specific implementation, the Global Mapper software also has the function of anti-aliasing images. Therefore, it can also directly call the Global Mapper software to perform anti-aliasing processing on each cropped terrain three-dimensional surface, so as to obtain a number of anti-aliased cropped terrain three-dimensional surfaces. Then, the three-dimensional surfaces processed as described above can be merged and smoothed, so as to complete the geometric correction of the triangular network terrain three-dimensional surface.

[0081] S53. Merge several antialiased and cropped 3D terrain surfaces to obtain a preprocessed triangular mesh terrain 3D surface. In specific applications, Global Mapper software can also be directly called to merge the 3D terrain surfaces after antialiasing to obtain the preprocessed triangular mesh terrain 3D surface. Finally, smooth the preprocessed triangular mesh terrain 3D surface to obtain the corrected triangular mesh terrain 3D surface. The smoothing process is as shown in step S54 below.

[0082] S54. Smooth the preprocessed triangular mesh terrain 3D surface to obtain the corrected triangular mesh terrain 3D surface after smoothing. In specific applications, but not limited to, first obtain the smoothing angle, and then smooth the preprocessed triangular mesh terrain 3D surface based on this smoothing angle. Further, the preprocessed triangular mesh terrain 3D surface can be imported into 3ds Max software, then call the smooth surface component in this software, and use the aforementioned smoothing angle as the smoothing parameter to smooth the preprocessed triangular mesh terrain 3D surface. In this way, the geometric correction of the triangular mesh terrain 3D surface can be completed, and a modifiable 3D surface with a more continuous and smoother surface geometric structure can be obtained. Optionally, for example, the corrected triangular mesh terrain 3D surface can be, but not limited to, a file in obj format.

[0083] After completing the geometric correction of the triangular mesh terrain 3D surface of the target to be constructed, the corrected triangular mesh terrain 3D surface can be used to adjust the 3D geometric structure of the aforementioned initial real-scene 3D model, thereby effectively removing the ribbon holes and independent holes in the model. Specifically, the model correction process can be, but not limited to, as shown in step S6 below.

[0084] S6. Use the corrected triangular mesh terrain 3D surface to perform geometric structure reconstruction processing on the initial real-scene 3D model to obtain the real-scene 3D model of the target to be constructed after the geometric structure reconstruction processing. In this embodiment, the geometric structure reconstruction process of the initial real-scene 3D model can be, but not limited to, as shown in steps S61 to S63 below.

[0085] S61. Based on the corrected triangular mesh terrain 3D surface, obtain the 3D geometric structure data of the target to be constructed.

[0086] S62. Use the 3D geometric structure data to adjust the geometric structure of the initial real-scene 3D model to obtain the preprocessed real-scene 3D model of the target to be constructed after the adjustment.

[0087] S63. Perform texture replacement processing on the preprocessed real-scene 3D model to obtain the real-scene 3D model of the target to be constructed after the texture replacement processing.

[0088] In this embodiment, the principles of the foregoing steps S61 to S63 are as follows: Use the corrected triangular network terrain three-dimensional surface to obtain the true three-dimensional geometric structure of the target to be constructed. Then, use this three-dimensional geometric structure to perform geometric correction on the initial real-scene three-dimensional model. Finally, perform texture replacement. After the replacement is completed, a real-scene three-dimensional model without ribbon holes and independent holes can be obtained. Optionally, but not limited to, import the corrected triangular network terrain three-dimensional surface into Smart 3D software, and then retain the texture of the initial real-scene three-dimensional model, and only use the corrected triangular network terrain three-dimensional surface to reset the geometric structure of the foregoing initial real-scene three-dimensional model. In this way, a preprocessed three-dimensional model with a more complete structure can be obtained. Based on the foregoing processing, the holes in the model can be effectively removed, and at the same time, models with real three-dimensional coordinates in different formats can be output, such as models in obj, osgb, fbx, 3D Tiles, 3MX and other formats. Thus, while improving the model accuracy, the scope of use of the model can be expanded. In addition, since the geometric structure of the initial real-scene three-dimensional model has changed, it is necessary to use the foregoing colored point cloud data to perform texture replacement on the initial real-scene three-dimensional model (i.e., the preprocessed real-scene three-dimensional model) with the adjusted geometric structure. After the replacement is completed, a complete real-scene three-dimensional model can be obtained. Of course, the foregoing texture replacement is also implemented based on Smart3D software.

[0089] Thus, through the method for constructing a real-scene three-dimensional model detailed in the foregoing steps S1 to S6, the present invention uses point cloud reverse modeling technology and geometric terrain surface correction technology to construct a real-scene three-dimensional model with high spatial accuracy. Therefore, the present invention effectively solves the problem of ribbon holes or independent holes existing in the real-scene three-dimensional model obtained by using point cloud reverse modeling technology, making the reconstructed real-scene three-dimensional model have a high utilization rate of the previous surveying and mapping results, and having the advantages of fast and simple operation, no holes in the model, high accuracy, high spatial resolution, good texture mapping effect, and multiple result data formats, and is suitable for large-scale application and promotion.

[0090] See Figures 2 to 6 As shown, in the second aspect of this embodiment, the method provided in the first aspect of the embodiment is used to perform real-scene three-dimensional modeling on the Yanmengou Basin in Area A of our country. The modeling process is as follows.

[0091] Taking the Yanmengou Basin in Area A as the research area, the method provided in the first aspect of the embodiment is used to perform DEM and DOM fusion on this research area to construct a terrain-level real-scene three-dimensional model.

[0092] The Yanmengou Basin is the first batch of demonstration basins with serious flash flood disasters selected in Area A. A standardized demonstration pilot project will be carried out in high-risk areas of flash flood disasters in 2022-2023 to enhance the ability to defend against flash flood disasters. The Yanmengou Basin is located on the eastern edge of the northwestern Sichuan Plateau. The terrain is high and slopes from southeast to northwest. The upper reaches of the basin are mountainous, with the top of the mountain above 4,000 meters above sea level. The confluence with the Minjiang River is about 1,360 meters above sea level, and the average altitude of the basin is about 2,500 meters. The basin is roughly fan-shaped, with a deep riverbed in the middle and upper reaches, turbulent water flow, large gradient, and a "V" shape in cross-section. The two banks are steep and composed of stones, sand and pebbles, and the river system is developed along the river, with many tributaries. The basin area is 113km 2 , is an ideal area for studying terrain-level real-life 3D models.

[0093] In this embodiment, the digital elevation data and digital orthophoto data of the Yanmengou area are first collected, and bilinear interpolation processing is performed on the digital elevation data. Then, gridded point cloud data with three-dimensional information is extracted from the interpolated data set. The point cloud attributes are shown in Table 1 below, and the point cloud extraction effect diagram is shown in Table 1 below. Figure 2 shown.

[0094] Table 1 is an attribute information table of point cloud data extracted without bilinear interpolation and point cloud data extracted after bilinear interpolation.

[0095] Table 1

[0096]

[0097] It can be concluded from the above Table 1 that after bilinear interpolation, the interval of the point cloud data in the digital elevation data changes from 2.4901m to 0.93m, the spacing becomes smaller, and the number increases; at the same time, the extracted point cloud data also increases from 5078724 to 146129500, and the number and density are higher. Therefore, compared with traditional point cloud data extraction, this embodiment can provide higher density point cloud data, so that more refined texture features can be obtained.

[0098] Next, this embodiment uses the method provided in the first aspect of the embodiment to colorize the aforementioned extracted point cloud data to obtain colored point cloud data; wherein, the effect diagram of the colored point cloud data can be seen in Figure 3 ;from Figure 3 It can be seen that after coloring, it is equivalent to adding the rendering of the real geographical environment, which facilitates the identification and application of land objects and achieves a realistic visualization effect.

[0099] Furthermore, in this embodiment, the colored point cloud data is imported into the Smart 3D software for 3D reconstruction, thereby obtaining an initial real-scene 3D model. However, the model has ribbon-type vulnerabilities and independent vulnerabilities, such as Figure 4As shown in the local area, further processing is required; that is, the digital elevation data of Yanmengou is used to generate its corresponding triangular network terrain three-dimensional surface, and geometric correction is performed on the triangular network terrain surface. The corrected effect diagram can be seen in Figure 5 as shown; from Figure 5 it can be seen that the geometric structure on its surface is more continuous and smoother.

[0100] Finally, the corrected triangular network terrain three-dimensional surface is imported into the Smart 3D software to adjust the three-dimensional geometric structure of the initial real-scene three-dimensional model of Yanmengou. Among them, the effect diagram of the real-scene three-dimensional model of Yanmengou obtained after reconstruction can be seen in Figure 6 as shown; from Figure 6 it can be seen that the model hole problem has been effectively solved, the texture is clear, and the requirements of the terrain-level real-scene three-dimensional model are met; at the same time, in this embodiment, according to the same-precision detection requirements in "Quality Inspection and Acceptance of Surveying and Mapping Results" GB / T 24356-2009, accuracy comparison and inspection are carried out with the original DEM. The accuracy comparison table is shown in Table 2 below.

[0101] Table 2 is the comparison table of the original DEN data and the accuracy of this model.

[0102] Table 2

[0103]

[0104]

[0105]

[0106]

[0107]

[0108] As can be seen from Table 2 above, a total of 85 inspection points are involved. The mean error in the plane is 0.0490m, and the mean error in elevation is 0.0263m. Therefore, the accuracy of the model constructed in this embodiment is consistent with the DEM accuracy; thus, it shows that the method provided in this embodiment can effectively solve the problems of ribbon holes and independent holes existing in traditional modeling, and can provide a terrain-level three-dimensional real-scene model with high accuracy, high spatial resolution, and good texture mapping effect, thereby providing a data model basis with reliable accuracy, clear texture, and diverse formats for mountain flood disaster analysis, which is of great significance for improving the mountain flood disaster prevention ability.

[0109] As Figure 7 shown, in the third aspect of this embodiment, a hardware device for implementing the method for constructing the real-scene three-dimensional model described in the first aspect of the embodiment is provided, including:

[0110] A data acquisition unit for acquiring model construction data of a target to be constructed, where the model construction data includes digital elevation data and digital orthophoto image data of the target to be constructed.

[0111] A point cloud processing unit for determining point cloud data for reconstructing a three-dimensional model of the target to be constructed based on the digital elevation data, and performing point cloud coloring processing on the point cloud data by using the digital orthophoto image data to obtain colored point cloud data.

[0112] A model reconstruction unit for constructing an initial real-scene three-dimensional model of the target to be constructed based on the colored point cloud data.

[0113] A surface construction unit for constructing a triangular network terrain three-dimensional surface of the target to be constructed by using the digital elevation data.

[0114] A geometric correction processing unit for performing geometric correction processing on the triangular network terrain three-dimensional surface to obtain a corrected triangular network terrain three-dimensional surface.

[0115] A model adjustment unit for performing geometric structure reconstruction processing on the initial real-scene three-dimensional model by using the corrected triangular network terrain three-dimensional surface, so as to obtain a real-scene three-dimensional model of the target to be constructed after the geometric structure reconstruction processing.

[0116] For the working process, working details and technical effects of the device provided in this embodiment, reference may be made to the first aspect of the embodiment, which will not be elaborated herein.

[0117] As Figure 8 shown, in the fourth aspect of this embodiment, another device for constructing a real-scene three-dimensional model is provided. Taking the device as an electronic device as an example, it includes: a memory, a processor, and a transceiver that are communicatively connected in sequence, where the memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program and execute the method for constructing a real-scene three-dimensional model as described in the first aspect of the embodiment.

[0118] Specifically, the memory may include, but is not limited to, random access memory (RAM), read only memory (ROM), flash memory, first input first output (FIFO), and / or first in last out (FILO), etc.; specifically, the processor may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor may be implemented in at least one of the following hardware forms: DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), and PLA (Programmable Logic Array). At the same time, the processor may also include a main processor and a coprocessor. The main processor is a processor used to process data in the wake state, also known as the CPU (Central Processing Unit); the coprocessor is a low-power processor used to process data in the standby state.

[0119] In some embodiments, the processor may be integrated with a GPU (Graphics Processing Unit), and the GPU is responsible for rendering and drawing the content to be displayed on the display screen. For example, the processor may be, but is not limited to, a microprocessor of the STM32F105 series, a reduced instruction set computer (RISC) microprocessor, an X86 architecture processor, or a processor integrated with a neural-network processing unit (NPU); the transceiver may be, but is not limited to, a Wi-Fi wireless transceiver, a Bluetooth wireless transceiver, a General Packet Radio Service (GPRS) wireless transceiver, a ZigBee (low-power local area network protocol based on the IEEE 802.15.4 standard) wireless transceiver, a 3G transceiver, a 4G transceiver, and / or a 5G transceiver, etc. In addition, the device may also include, but is not limited to, a power module, a display screen, and other necessary components.

[0120] For the working process, working details, and technical effects of the electronic device provided in this embodiment, reference may be made to the first aspect of the embodiment, which will not be elaborated here.

[0121] The fifth aspect of this embodiment provides a storage medium storing instructions for the method of constructing the real-scene three-dimensional model described in the first aspect of the embodiment, that is, instructions are stored on the storage medium, and when the instructions run on a computer, the method of constructing the real-scene three-dimensional model described in the first aspect is executed.

[0122] Among them, the storage medium refers to a carrier for storing data, and can include, but is not limited to, floppy disks, optical discs, hard disks, flash memories, USB flash drives, and / or Memory Sticks, etc. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.

[0123] For the working process, working details, and technical effects of the storage medium provided in this embodiment, reference can be made to the first aspect of the embodiment, which will not be elaborated here.

[0124] The sixth aspect of this embodiment provides a computer program product containing instructions, which, when running on a computer, cause the computer to execute the method of constructing the real-scene three-dimensional model described in the first aspect of the embodiment. Among them, the computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.

[0125] Finally, it should be noted that the above are only preferred embodiments of the present invention and are not used to limit the protection scope of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for constructing a real - scene three - dimensional model, characterized in that, Including: Obtain the model construction data of the target to be constructed, where the model construction data includes the digital elevation data and digital orthophoto image data of the target to be constructed; Based on the digital elevation data, determine the point cloud data for reconstructing the three-dimensional model of the target to be constructed, and use the digital orthophoto image data to perform point cloud coloring processing on the point cloud data to obtain colored point cloud data; Based on the colored point cloud data, construct the initial real scene three-dimensional model of the target to be constructed; Use the digital elevation data to construct the triangular network terrain three-dimensional surface of the target to be constructed; Perform geometric correction processing on the triangular network terrain three-dimensional surface to obtain the corrected triangular network terrain three-dimensional surface; Use the corrected triangular network terrain three-dimensional surface to perform geometric structure reconstruction processing on the initial real scene three-dimensional model, so as to obtain the real scene three-dimensional model of the target to be constructed after the geometric structure reconstruction processing.

2. The method according to claim 1, characterized in that, Based on the digital elevation data, determining the point cloud data for reconstructing the three-dimensional model of the target to be constructed includes: Use the bilinear interpolation algorithm to perform bilinear interpolation processing on the digital elevation data to obtain an interpolated digital elevation data set; Determine the grid point cloud data with three-dimensional coordinate information from the interpolated digital elevation data set, and use the grid point cloud data as the point cloud data for reconstructing the three-dimensional model of the target to be constructed.

3. The method according to claim 2, wherein Determining the grid point cloud data with three-dimensional coordinate information from the interpolated digital elevation data set includes: Call the pixel center point extraction component in Global Mapper to perform grid point cloud extraction processing on the interpolated digital elevation data set, so as to obtain the grid point cloud data with three-dimensional coordinate information after the grid point cloud extraction processing.

4. The method according to claim 1, wherein Using the digital orthophoto image data to perform point cloud coloring processing on the point cloud data to obtain colored point cloud data includes: Call the pixel matching component in Global Mapper to use the pixel matching component to match each pixel data in the digital orthophoto image data with the point cloud data, so as to determine the pixel data matching each point cloud data in the digital orthophoto image data; For any point cloud data, obtain the color information of the pixel data corresponding to the any point cloud data as the coloring information; Add the coloring information to the any point cloud data, so as to complete the point cloud coloring processing of the any point cloud data after the addition.

5. The method according to claim 1, characterized in that Performing geometric correction processing on the triangular network terrain three-dimensional surface to obtain the corrected triangular network terrain three-dimensional surface includes: Perform cropping processing on the triangular network terrain three-dimensional surface to obtain several cropped terrain three-dimensional surfaces; Perform anti-aliasing processing on each cropped terrain three-dimensional surface in the several cropped terrain three-dimensional surfaces to obtain several anti-aliased cropped terrain three-dimensional surfaces; Merge the several anti-aliased cropped terrain three-dimensional surfaces to obtain a preprocessed triangular network terrain three-dimensional surface; Smooth the preprocessed triangular mesh terrain three-dimensional surface so as to obtain the corrected triangular mesh terrain three-dimensional surface after the smoothing process.

6. The method according to claim 5, wherein Smoothing the preprocessed triangular mesh terrain three-dimensional surface so as to obtain the corrected triangular mesh terrain three-dimensional surface after the smoothing process includes: Obtain a smoothing angle, and based on the smoothing angle, smooth the preprocessed triangular mesh terrain three-dimensional surface to obtain the corrected triangular mesh terrain three-dimensional surface.

7. The method according to claim 1, wherein Use the corrected triangular mesh terrain three-dimensional surface to perform geometric structure reconstruction processing on the initial real-scene three-dimensional model, including: Based on the corrected triangular mesh terrain three-dimensional surface, obtain the three-dimensional geometric structure data of the target to be constructed; Use the three-dimensional geometric structure data to adjust the geometric structure of the initial real-scene three-dimensional model so as to obtain the preprocessed real-scene three-dimensional model of the target to be constructed after the adjustment; Perform texture replacement processing on the preprocessed real-scene three-dimensional model so as to obtain the real-scene three-dimensional model of the target to be constructed after the texture replacement processing.

8. A construction device for a real - scene three - dimensional model, characterized in that, Includes: A data acquisition unit for acquiring model construction data of the target to be constructed, where the model construction data includes the digital elevation data and digital orthophoto image data of the target to be constructed; A point cloud processing unit for determining point cloud data for reconstructing the three-dimensional model of the target to be constructed based on the digital elevation data, and performing point cloud coloring processing on the point cloud data using the digital orthophoto image data to obtain colored point cloud data; A model reconstruction unit for constructing an initial real-scene three-dimensional model of the target to be constructed based on the colored point cloud data; A surface construction unit for constructing a triangular mesh terrain three-dimensional surface of the target to be constructed using the digital elevation data; A geometric correction processing unit for performing geometric correction processing on the triangular mesh terrain three-dimensional surface to obtain a corrected triangular mesh terrain three-dimensional surface; A model adjustment unit for using the corrected triangular mesh terrain three-dimensional surface to perform geometric structure reconstruction processing on the initial real-scene three-dimensional model so as to obtain the real-scene three-dimensional model of the target to be constructed after the geometric structure reconstruction processing.

9. An electronic device, characterized in that, Includes: A memory, a processor, and a transceiver that are communicatively connected in sequence, where the memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program and execute the method for constructing a real-scene three-dimensional model according to any one of claims 1 to 7.

10. A storage medium, characterized in that, Instructions are stored on the storage medium, and when the instructions are run on a computer, the method for constructing a real-scene three-dimensional model according to any one of claims 1 to 7 is executed.

Images